Dynamic road salt model at watershed scale

In the Jefferson Project, we have detailed salt loading measurements made with IoT devices disposed in the municipal salt trucks used at roads crossing forest or small cities around lake George (NY-US). In some streams we measure conductivity and flow, what allows the evaluation of Chloride concentration and mass balance estimation between the input from roads and the amount flowing through streams to the lake. To estimate the concentrations to all the other streams not measured, we established a WRF-hydro simulation that incorporates weather+ land surface model + a simple groundwater element and incorporates energy and mass balance. With the aim of creating the simplest model to provide mass transfer, we propose to use some of its parameters, related to the snow melting of the WRF-H model, as input in a new code. This code simulates, at watershed scale (and so is not a gridded representation as in WRF-H), the stream concentrations. The code defines the watersheds at positions where there is a sensor available and incorporates the timing of road salt deposition and melting at each of these watersheds into a group of delay functions that are compared with measured concentrations.

The goal of this study is: 1) Based on the dataset, discuss the physical interpretation of the system behavior (i.e. the higher concentrations occur after the winter indicating that the groundwater stores most of the salt mass). 2)Discuss intermediate results, difficulties/pitfalls of the dynamic salt model created, when comparing with measurements in the scale of the watershed. The result of this study shows the relative importance of each transfer compartment in the overall process of Chloride mass transfer at this watershed scale.